The retinoblastoma (RB) gene is a potent suppressor of human cancers. While RB is thought to have multiple functions during the cell cycle, the specific functions of RB that are critical for its tumor suppression activity are still unclear. We propose to test the idea that RB and its two related family members p107 and p130 play a pivotal role in preventing cancer initiation in vivo by controlling the transition between the GO and G1 phase of the cell cycle. Specifically, we hypothesize that the acute somatic loss of RB family function leads to defects in both cell cycle exit (G1->GO) and cell cycle re-entry (GO->G1) in critical cell populations, and that these defects may result in the initiation of tumorigenesis. We will test our hypothesis using mouse models with combinations of germline and conditional mutations of RB family genes. This mouse genetic approach will enable us to determine the role of RB family genes in spatially and temporally-defined cell populations. These mouse models will also allow us to recapitulate in vivo the early stages of cancer in humans. Studies in the first Specific Aim will characterize the function of RB family genes in cell cycle exit during embryonic development. RB patients develop retinoblastomas, which are retinal neural tumors, as early as the fetal stage. We will focus our studies on neural progenitors to explore the consequences of loss of RB family function on cell cycle exit and differentiation defects leading to cancer initiation. Studies in the second Specific Aim will determine if the RB family proteins are required for maintenance of differentiation in vivo. Specifically, we will investigate the consequences of acute loss of function of RB family genes on cell cycle re-entry of differentiated hepatocytes in vivo. These studies will be important to determine if some terminally differentiated cells may be at the origin of cancer. Studies in the third Specific Aim will combine molecular approaches and RNA interference in vivo to understand the molecular mechanisms underlying cell cycle re-entry upon loss of RB family function in vivo. We will focus our studies on candidate mediators of cell cycle re-entry in RB family mutant hepatocytes, including the E2F transcriptional activator and the Id2 transcriptional repressor. Understanding the molecular bases and the cellular consequences of the loss of RB family tumor suppressors in vivo will give important and novel insights into the mechanisms of cancer initiation. These studies will further provide the foundation for the detection and treatment of early stages of human cancer.